MerlinSafety/Pluto
Pluto is a 9B parameter coding and reasoning model developed by Merlin Research, built on Qwen/Qwen3.5-9B-Base. It features a precision-first design, a massive 1,000,000 token context window, and incorporates Adaptive Entropy Regularization with quantum noise from IBM Quantum hardware. Pluto is specifically optimized for agentic coding environments and complex technical reasoning tasks, excelling in code generation and analysis for large codebases.
Loading preview...
Pluto: Precision Coding and Reasoning Model
Pluto is a 9 billion parameter model from Merlin Research, engineered for precision, robustness, and seamless deployment in agentic coding environments. Built upon Qwen/Qwen3.5-9B-Base, its training prioritizes error minimization over fluency, making it highly effective for critical coding tasks.
Key Capabilities & Features
- 1M Token Context: Handles extensive codebases and long conversation histories without chunking, maintaining coherent reasoning across the full context window.
- Agentic Deployment Ready: Fine-tuned for integration with Claude Code, OpenAI Codex/Assistants API, and local deployment workflows (GGUF/quantized variants).
- Quantum Entropy Regularization (AER): Utilizes quantum noise from IBM Quantum Kingston during RL training to enhance robustness, prevent entropy collapse, and improve stability on out-of-distribution inputs.
- Distillation from Frontier Models: Incorporates knowledge from advanced coding models and a private dataset of reasoning traces to achieve deep reasoning at a 9B scale.
Ideal Use Cases
- Complex code generation and refactoring.
- Multi-file codebase analysis and code review.
- Agentic coding pipelines and technical reasoning.
- Local deployment for large private codebases.